1

Introduction to LBW for ALFALFA followup

Martha Haynes UAT13 13.01.14

ALFALFA source codes

2

ALFALFA HI detections are coded according to:

Code 1 High quality sources, typically with S/N > 6.5

Code 2 Sources of lower S/N which are coincident with a probable OC of the same redshift (known from another source) => the “priors”

Code 3 Low S/N sources without identifiable OC

Code 4 Low S/N sources with a possible OC of unknown redshift

Code 5 Corresponding to Code 2, but of such low S/N or possible RFI contamination that they are untrustworthy

Code 6 Like OH megamasers at 0.16 < cz < 0.24

Signal Extractor -- Introduction ● The signals are extracted by cross-correlations of a

template with the spectra. => “matched-filter algorithm

• More sensitive than peak-finding algorithms.

• Sensitive to total flux, not only peak flux

• Especially important for low mass systems

• Using FFT's, cross-correlations are computationally fast

Saintonge 2007 AJ 133, 2087

Slide: Amelie Saintonge

Saintonge 2007 AJ

133, 2087

Signal Extractor -- Application(2)

The process is :

• Repeat for a range of widths of the template

• e.g. 10 km/s – 600km/s

• Choose the width for which the convolution is maximised --> position of the signal

• Calculate the amplitude of the signal from the width

Slide: Amelie Saintonge

High S/N

5

Low S/N

6

“Dark” galaxy candidate

7

Identifying Optical Counterparts

ALFALFA source centroids good to ~18” (depends on S/N) ALFALFA catalogs include: • the HI centroid position • the position of the most

probable OC • OC’s SDSS PhotoObjID and

SpecObjID (where applicable)

Of 15855 sources in α.40: • 1013 have no OC • 844 of those could be HVCs

(or LG minihalos) • 199 (<2%) extragalactic • Of those, <50 are “isolated”

Suppose HIPASS detects a source at S/N~6 near 3000 km/s in this field. The position error box will have a radius of ~2.5’.

The opt counterpart could be gal #1, 2, 3, 4, 5 or 6.

ALFALFA will detect the same source with S/N~50

Centroiding accuracy goes roughly as HPFW(PSF)/(S/N)

and the Arecibo beam is ¼ as wide as the Parkes one

The same source will have an ALFALFA position error of ~ 0.1’

1

2

3

4

5

6

ALFALFA advantage for finding the OC

Dark galaxies

• In agreement with previous results, ALFALFA finds that fewer that 2% of (clearly extragalactic) HI sources cannot be identified with an optical counterpart.

• The majority of objects without OC’s are found near to galaxies with similar redshifts.

• There are few interesting cases to be confirmed (work in progress).

“Dark” object in a group

ALFALFA HI on SDSS r

HI peak with no optical/marginal UV: almost dark?

MH, RG (CU), Cannon (Macalester), Salzer (Indiana), Karen Lee-Waddell, K. Spekkens

(RMCC/Queens)

Ongoing deeper B,R band imaging of HI

detections with no OC

Salzer, Adams

13

• Arise in interacting/merging galaxy systems. • When galaxies merge, gas clouds close to

their nuclei are shocked and heated by the collision, and the emission from certain molecules especially OH is strongly amplified.

• Since this stimulated emission is like the more familiar laser but occurs in the microwave region of the electromagnetic spectrum, it is called a "maser".

• When galaxies collide, the emission is millions of times stronger than in normal galaxies, hence the term "megamaser".

• Such objects are also typically (ultra) luminous in the far-infrared.

OH Megamasers: OHMs

14

18 cm OHMs

18 cm OH lines

15

• First molecular lines detected in radio.

• 18 cm lines formed by “lambda doubling” hyperfine structure

• 1612, 1665, 1667 and 172O MHz

Pacholczyk, Radio Astrophysics

16

Redshifted 18 cm OHMs in ALFALFA • f5 = 1665.4018 • f7 = 1667.3590 In OHMs, S(f7)>S(f5) fobs = frest / (1+z) ALFALFA: 1340-1430 MHz, corresponding, for OH, to 0.166 < z < 0.244

• Emission at f > 1422 MHz (blueshifted if HI) • Emission associated with OC in 0.1666 < cz < 0.244 • Emission with no OC

17

OHM candidates in α.40

AGC 181310: • Previously discovered OHM by

Darling & Giovanelli (2001) • Coincides with SDSS

J083212.61+275139.8 at z=0.167830

• Also IRAS 08201+2801 and 5C 07.206

We need to confirm the others, and a few more low SNR sources.

18

AGC 181310: confirmed OHM

cz(HI) = -1518 km/s

ALFALFA sensitivity & completeness

19

• We want to integrate longer on the low S/N sources.

• Even on high S/N sources, we want to verify they are real.

• Point at OC if there is one or the HI centroid if not

“Targeted observations”

• LBW has a single horn (“pixel”) but higher gain and lower Tsys

than ALFA

Position-switched observing

20

The signals we are trying to take are billions of times weaker than the radio noise contributed by the receiver, electronics, antenna, cosmic microwave background and the sky overall. Somehow, we have to subtract off all those unwanted contributions to find our signal. We assume that a random position in the sky does not contain an HI line source at the exact same velocity as our target source. We observe such a position, but track it over the exact same Az, ZA as our observations of the target source. => ON-OFF pair

Position-switched observing

21

• Position telescope on target source • Track ON-source for 3 minutes • Move to same Az,ZA as at start of ON-

source track but 4 mins from now; this is the OFF-source position.

• Track OFF-source for 3 minutes. • Record noise with “CAL” (noise diode) –

ON for 10 secs; then record noise for 10 sec with CAL-OFF.

• Go to next target. • Repeat ON-OFF-CAL ON/OFF sequence

This is what the “command file” (for a set of sources for the whole night) does.

22

Beam 0

Beam 1

Beam 2

Beam 3

Beam 4

Beam 5

Beam 6

WAPP

ALFA spectra: 16 x 4096 frequency channels (2 not used) 7 beams X 2 polarizations/beam 100 MHz wide Centered at 1385 MHz

Spectrometer setup for LBW

So resolution is 100 MHz/4096

channels

23

LBW

Interim Correlator

LBW spectra: 4 x 2048 frequency channels 1 beam X 2 polarizations/beam Low res/high res 25 MHz / 6.25 MH Centered at doppler velocity of target

Spectrometer setup for LBW

So resolution is

Low res: 25 MHz/2048 channels

High res: 6.25 MHz/2048 channels

pol A

pol B

Low res

High res

Estimating how long we integrate

24

The radiometer equation for our observations Δf : is the bandwidth per channel ts : is the effective integration time, in secs ft : accounts for the degree of smoothing, the technique

applied for bandpass subtraction, clipping losses, etc. The factor of 2 under the square root comes from the fact

that we average the two independent polarizations. See Giovanelli + 2005, AJ 130, 2598

For LBW, T~ 30K

G ~ 11 K/Jy

A2669/A2707 LBW followup

25

Targeted LBW observations of selected ALFALFA sources:

1. “Dark” galaxy candidates: high quality (Code 1) detections with no OC and not associated with known tidal debris fields

2. OH megamaser (OHMs) candidates: either at large blueshift or coincident (within centroiding accuracy) of OC of appropriate cz.

3. “Low mass dwarf candidates”: low signal-to-noise ratio sources at low cz (< 1000 km/s)

4. Statistical samples of low S/N signals possibly associated with optical galaxies (A2669 only)

ALFALFA: effective integration time of 40 seconds/beam LBW: 3 minutes ON-source

Scavenger Hunt #1

26

Next: 15 min break Then: Work on SH #1

Work in your groups ASK questions!

15:15: Remote observing demo